How to Design a Deep Learning Neural Network ?

Welcome back readers !!! Hope you all are doing well 😊.

In today's article, we are going to learn about the designing of deep learning neural network.

Just like we need a number of ingredients to prepare some food, there are certain ingredients to design a neural network. The main requirements to design a neural network are to have an appropriate arrangement of deep neural network layers, training parameters for hyperparameter tuning, and obviously the selection of appropriate dataset (or shall I say the set of images) on which the training is to be performed.

List of Layers in Deep Neural Network

  1. Input Layer
  2. Convolution Layer and its parameters
  3. LSTM Layer
  4. GRU Layer
  5. Fully Connected Layer
  6. Flatten Layer
  7. Normalization Layer
  8. Pooling Layer and its types
  9. Activation Layer
  10. Combination Layer
  11. Output Layer

Training Parameters in Deep Learning

  1. Optimizers
  2. Monitoring Progress
  3. Mini-batch Options
  4. Validation
  5. Solver Options
  6. Gradient Clipping

Selection of Appropriate Dataset

These individual headings will be discussed in detail in our upcoming articles. So stay tuned and keep supporting 😊. Kindly give your valuable suggestions in the comments section 🙏.

Akshay Juneja authored 15+ articles for INFO4EEE Website on Deep Learning.

No comments

Powered by Blogger.